The taxonomic data used in this website were obtained from eggs and larvae collected during ichthyoplankton surveys conducted from 1972 to 2015 by the Recruitment Processes Program (Table 1). Data on distribution and abundance of eggs and larvae were based on surveys from 1972-2015, excluding 1973-1976. Collection data for cruises conducted up to 1988 can be found in Dunn and Rugen (1989), and in the AFSC ichthyoplankton cruise database (Rugen, 2008) for those cruises conducted from 1989 to 2015. The majority of the data are from samples collected using a MARMAP type bongo sampler (Posgay and Marak, 1980) with an inside diameter of 60 cm and a 0.333 or 0.505-mm mesh net. Before 1985, standard MARMAP oblique tows were made to 200 m following MARMAP sampling procedures (Smith and Richardson 1977). In 1985, sampling depth in the Gulf of Alaska was changed to near bottom in order to accurately determine the abundance of Walleye Pollock eggs and early larvae. Beginning in the early 1990s, annual larval surveys in the Gulf of Alaska began to sample to 100 m in late May because that is where the larvae are most abundant. Flowmeters suspended in the mouths of the nets of all ichthyoplankton gear were used to determine the volume of water filtered by each net. Data from 1-m Tucker trawls were used only for cruises in which Tucker trawls were the primary gear (14 cruises, 1035 tows; Table 1).
A Sameoto neuston sampler (Sameoto and Jaroszyinski, 1969), with a mouth opening 0.3 m high x 0.5 m wide and a 0.505-mm mesh net, was used sporadically throughout the time series to collect eggs and larvae that reside in the upper surface waters. A partial summary and analysis of our spring neuston collections from the Gulf of Alaska (1981-1986) are presented by Doyle et al. (1995). A summary and analysis of our neuston collections off the U.S. west coast are presented by Doyle (1992). In comparison to bongo collections presented here (Table 1), Table 2 presents an overall summary, based on the standard data set for 1972-1996 (Matarese et al., 2003), of the 20 most common taxa collected with neuston gear and arranged by percent frequency of occurrence (Table 2). Data from the neuston tows were selected to generate maps for 15 taxa for which the best geographic distribution pattern was described using surface gear (Table 3).
The sampled population is defined as fishes in the size range that is effectively caught by the sampling gear, and may include both larvae and juveniles because the transformation point between these two stages is unknown for many species, and because fishes were staged inconsistently over the years. The number of individuals caught in each of the three sampling gears (i.e., bongo net, neuston net, and Tucker trawl) was standardized to number caught per 10 m2 of surface area. Catches from bongo and Tucker gear were standardized based on net mouth area and tow depth and length (Smith and Richardson, 1977). Some of the Tucker tows were depth stratified; that is, two nets sampled two contiguous depth intervals. In these cases, the catches per 10 m2 from both nets were summed to integrate over the depths sampled by both nets. Since both gears were fished at similar depth ranges, we assumed that bongo and Tucker gear sample essentially the same population, and thus allowed data from these two gears to be combined. This assumption is supported by Shima and Bailey (1994), who concluded that the fish-length-specific sampling effectiveness of these two gears is not significantly different. It was also assumed that no individuals occurred below the depth sampled by the gear, thus the number caught per 10 m2 of surface area represents the total number of individuals in the water column below a surface area of 10 m2. Samples collected by neuston gear were standardized to number caught per 10 m2 of surface area based on net mouth width and tow length. Neuston data are usually represented as number per 1000 m3, but we chose to scale the data to surface area because that describes the number of animals in a specified area of the neuston layer. In comparing neuston catches and bongo catches per unit area, the neuston catches are much smaller because the volume of water filtered is much smaller. The neuston net effectively samples the top 15 cm of the surface layer; hence animals occurring below this depth are not sampled. It was assumed that most individuals of neustonic species occur in the top 15 cm, although many species migrate vertically in and out of the neuston.
Plankton samples were preserved in the field using a 5% formalin-seawater solution buffered with calcium carbonate chips or sodium borate; after 1983, fish larvae were transferred to 70% ethanol after formalin fixation. All fish eggs, larvae, and juveniles have been removed and identified to the lowest possible taxon since 1980 at the Plankton Sorting and Identification Center in Szczecin, Poland (Fig. 1). Identifications are verified by the taxonomic team at AFSC using information found in Matarese et al. (1989) and supplemented by a number of more recent publications including those by Moser (1996), Busby (1998), Orr and Matarese (2000), Blood and Matarese (2010), and Matarese et al. (2013).
The study area extends from the Arctic (Chukchi and Beaufort seas) through the Bering Sea, into the Gulf of Alaska, and along the U.S. west coast (Map 9). Most of the sampling throughout the time series occurred in Shelikof Strait and west toward the sea valley and along the Alaskan Peninsula (Map 10). Repeated sampling was more extensive with bongo/Tucker gear than with neuston gear (Map 11 and Map 12). Coverage is most complete in the Gulf of Alaska (except for 1980 when no sampling occurred there; Fig. 2) and less extensive in the Bering Sea and off the U.S. west coast. Coverage is also more complete along shelf regions and less extensive in deeper ocean waters where mesopelagic and deepwater flatfishes spawn. No ichthyoplankton sampling occurred in the North Pacific in Canadian waters. Most of the 625 km2 grid cells were sampled 1-10 times.
Most of the data used in this study were obtained from bongo tows taken during April and May. Combined bongo tows and Tucker trawls during these two months account for 52-79% of the yearly total, depending on geographic region (Fig. 3). Data from neuston tows were more evenly distributed throughout the year, with tows made during April and May accounting for 0-54% of the yearly total in each region (Fig. 3). The distribution of combined bongo tows/Tucker trawls and neuston tows by year and geographic region further illustrates our extensive coverage of the Gulf of Alaska during the 11-yr period (1977-1987) of routine sampling with neuston gear (Fig.4).
The increase in our taxonomic knowledge over 40 years has allowed our basic knowledge of the early life history of species to be expanded and fine-tuned. Of the 646 known fish species that occur in the Northeast Pacific Ocean and Bering Sea, we can currently identify a portion of the early life history stages for about 353 (Deary et al. in prep.). Phylogenetic order of higher taxa generally follows Nelson (2007); genera and species are listed alphabetically within families.
Larvae -- Most of the larvae included in the IIS can be found in our laboratory guide (Matarese et al., 1989), atlas (Matarese et al., 2003), or CalCOFI Atlas 33 (Moser, 1996). Any taxa that spawn in our study area and have been described with accompanying illustrations can be found within this website. Taxa that do not produce planktonic early life history stages are generally excluded (e.g., Embiotocidae) as are strictly freshwater and estuarine species or spawners. Sources useful in identifying early life history stages of freshwater and estuarine species found adjacent to our study area include Wang (1981, 1986) and Auer (1982). Updated and new ELH data and figures will be added as information is published.
The complete list of larval fish taxa collected from Recruitment Processes Program cruises 1972 through 1996 used for initial consideration in the atlas (Matarese et al., 2003) is presented in Table 4. This list was reduced to 102 taxa within 34 families (Table 5). All individual species covered in Matarese et al. (2003) are found in this website; taxa covered only at the genus or family level are not included. Additional maps are added as distributional data are published. Currently, 296 taxa are included in this website; distribution maps are available for over 110 taxa.
Eggs -- Similar to the larvae as described above, most of the fish egg taxa included in the IIS can be found in our laboratory guide (Matarese et al., 1989), atlas (Matarese et al., 2003), or CalCOFI Atlas 33 (Moser, 1996). Of larval taxa in the IIS, any of their respective eggs that have been described can be found within this website; most/many have accompanying illustrations.
The complete list of fish egg taxa collected from Recruitment Processes Program cruises 1972 through 1996 used for initial consideration in the atlas is presented in Table 6. Data were restricted to pelagic fish eggs due to sampling strategies and gear limitations. The selection process resulted in a total of 30 taxa included in 14 families (Table 7). For each of the 28 individual species in the final list, distributional maps depicting presence/absence are presented in the IIS. As with larvae, additional egg distribution maps are added as data are published. Currently, distribution maps are available for 29 taxa.
General life history -- Life history data are provided as ancillary information, which may aid in identification of eggs and larvae. These data were extracted from the current general literature as well as original unpublished material. Geographic ranges are obtained from three types of sources: RACEBASE, an oracle database developed by the Resource Assessment and Conservation Engineering Division (RACE) which contains information collected by fisheries scientists on research survey cruises; ICHBASE, a database developed by AFSC that contains information on larval fish and eggs collected from research cruises; and from current literature. Mecklenburg et al., 2002 provided a significant amount of new information. Range information is restricted to the study area. Thus, the limits of the southern distribution beyond the California-Mexican border (SSC = South of southern California, below 32° 30' N), the northern range beyond the Arctic to the north and east, and the western range beyond the Bering Sea are not specified.